U.S. patent number 7,310,119 [Application Number 10/862,685] was granted by the patent office on 2007-12-18 for adaptive circuit for y-c separation.
This patent grant is currently assigned to Cirrus Logic, Inc.. Invention is credited to James A. Antone, Daniel O. Gudmundson, Rahul Singh.
United States Patent |
7,310,119 |
Singh , et al. |
December 18, 2007 |
Adaptive circuit for Y-C separation
Abstract
An adaptive circuit and method for separating luminance and
chrominance components from a composite video signal by deriving
three input lines from the composite video signal, determining
whether any luminance similarity exists among the three input
lines, and then selectively enabling a component filter based on
any luminance similarity. If no luminance similarity exists among
all three of the input lines, then a subtractive comb filter is
enabled to maintain high vertical luminance resolution. If
luminance similarity exists among all three of the input lines,
then an additive comb filter is enabled. The additive comb filter
performs three-line averaging when a high degree of similarity
exists among all three consecutive input lines to minimize
cross-chroma artifacts on lines that are similar. Chrominance
similarity among the three input lines can also be determined by
generating first and second chrominance values using different
pairs of the three input lines, computing a difference of the first
and second chrominance values, and comparing the chroma difference
to a threshold value. If no luminance similarity exists between any
of the three input lines or no chrominance similarity exists (i.e.,
there is vertical chroma transition), then a notch filter
incorporated into the subtractive comb filter can be enabled.
Inventors: |
Singh; Rahul (Austin, TX),
Gudmundson; Daniel O. (Dripping Springs, TX), Antone; James
A. (Austin, TX) |
Assignee: |
Cirrus Logic, Inc. (Austin,
TX)
|
Family
ID: |
38825946 |
Appl.
No.: |
10/862,685 |
Filed: |
June 7, 2004 |
Current U.S.
Class: |
348/666; 348/667;
348/E9.036 |
Current CPC
Class: |
H04N
9/78 (20130101) |
Current International
Class: |
H04N
9/78 (20060101) |
Field of
Search: |
;348/663-670 |
References Cited
[Referenced By]
U.S. Patent Documents
Primary Examiner: Lee; M.
Attorney, Agent or Firm: Lin; Steven Musgrove; Jack V.
Claims
What is claimed is:
1. A method of separating luminance and chrominance components from
a composite video signal, comprising: deriving at least three
consecutive input lines from a composite video signal; determining
whether any similarity exists among the three consecutive input
lines; and selectively enabling one of a plurality of component
filters based on said determining to separate the luminance and
chrominance components, wherein an additive comb filter is enabled
when luminance similarity exists among all three consecutive input
lines.
2. The method of claim 1 wherein determining whether any similarity
exists further comprises determining whether any luminance
similarity exists among the three consecutive input lines.
3. The method of claim 2 wherein determining whether any luminance
similarity exists further comprises: generating a first luminance
difference between a first one of the consecutive input lines and a
second one of the consecutive input lines; generating a second
luminance difference between the second consecutive input line and
a third one of the consecutive input lines; and comparing each of
the first and second luminance differences to a threshold
value.
4. The method of claim 2 wherein determining whether any similarity
exists further comprises determining whether any chrominance
similarity exists among the three consecutive input lines.
5. The method of claim 4 wherein determining whether any
chrominance similarity exists further comprises: generating a first
chrominance value using a first one of the consecutive input lines
and a second one of the consecutive input lines; generating a
second chrominance value using the second consecutive input lines
and a third one of the consecutive input lines; computing a
difference of the first and second chrominance values; and
comparing the difference to a threshold value.
6. The method of claim 1 wherein said additive comb filter performs
three-line averaging when a high degree of similarity exists among
all three consecutive input lines.
7. The method of claim 1 wherein selectively enabling one of the
plurality of component filters further comprises enabling a
subtractive comb filter when no luminance similarity exists among
all three consecutive input lines.
8. The method of claim 1 wherein selectively enabling one of the
plurality of component filters further comprises enabling a notch
filter when no luminance similarity exists between any of the three
consecutive input lines.
9. The method of claim 1 wherein selectively enabling one of the
plurality of component filters further comprises enabling a notch
filter when no chrominance similarity exists between the three
consecutive input lines.
10. A method of separating luminance and chrominance components
from a composite video signal, comprising: deriving a plurality of
consecutive input lines from a composite video signal; determining
whether any similarity exists between the consecutive input lines;
and based on said determining, selectively enabling one filter
among a notch filter, an additive comb filter, and a subtractive
comb filter to separate the luminance and chrominance
components.
11. The method of claim 10 wherein: deriving the plurality of
consecutive input lines further comprises deriving at least three
consecutive input lines from the composite video signal; and
determining whether any similarity exists further comprises
determining whether any similarity exists among the three
consecutive input lines.
12. The method of claim 11 wherein determining whether any
similarity exists further comprises determining whether any
luminance similarity exists among the three consecutive input
lines.
13. The method of claim 11 wherein determining whether any
similarity exists further comprises determining whether any
chrominance similarity exists among the consecutive input
lines.
14. The method of claim 13 wherein selectively enabling one of the
notch filter, the additive comb filter, or the subtractive comb
filter further comprises enabling the additive comb filter when
luminance similarity exists among all three consecutive input
lines.
15. The method of claim 13 wherein selectively enabling one of the
notch filter, the additive comb filter, or the subtractive comb
filter further comprises enabling the subtractive comb filter when
no luminance similarity exists among all three consecutive input
lines.
16. The method of claim 11 wherein selectively enabling one of the
notch filter, the additive comb filter, or the subtractive comb
filter further comprises enabling the notch filter when no
luminance similarity exists between the three consecutive input
lines and no chrominance similarity exists between the three
consecutive input lines.
17. An adaptive decision unit for controlling an electronic video
decoder having a plurality of component filters which separate
luminance and chrominance components from a composite video signal,
the adaptive decision unit comprising: first, second, and third
input lines; an adder circuit which generates first and second
luminance differences based on said first, second, and third input
lines, said adder circuit including first, second, and third notch
filters connected respectively to said first, second and third
input lines which produce first, second, and third notch-filtered
luminance signals, a first invert adder having a negative input
which receives the first notch-filtered luminance signal and a
positive input which receives the second notch-filtered luminance
signal, wherein the first invert adder generates the first
luminance difference, and a second invert adder having another
negative input which receives the third notch-filtered luminance
signal and another positive input which receives the second
notch-filtered luminance signal, wherein the second invert adder
generates the second luminance difference; and a logic circuit
which receives the first and second luminance differences and
activates one or more output signals indicative of any luminance
similarity among the first, second, and third input lines.
18. The adaptive decision unit of claim 17 wherein said logic
circuit determines the luminance similarity by comparing the first
and second luminance differences to a threshold value.
19. The adaptive decision unit of claim 18 wherein the threshold
value is a programmably set value.
20. The adaptive decision unit of claim 17 wherein said logic
circuit activates one of the one or more output signals to indicate
that all three of said first, second, and third lines have some
luminance similarity.
21. The adaptive decision unit of claim 20 wherein said logic
circuit activates another one of the one or more output signals to
indicate that all three of said first, second, and third lines have
a high luminance similarity.
22. The adaptive decision unit of claim 20 wherein said logic
circuit activates another one of the one or more output signals to
indicate that at least two of said first, second, and third lines
have a high luminance similarity.
23. The adaptive decision unit of claim 17, further comprising a
filter circuit which generates first and second chrominance signals
based on said first, second, and third input lines, wherein said
logic circuit receives the first and second chrominance signals and
activates one of the output signals to indicate that no luminance
similarity exists among the first, second, and third input lines
and that said first, second, and third lines have chrominance
similarity.
24. The adaptive decision unit of claim 23 wherein said filter
circuit includes: a first invert adder having a negative input
which receives the first input line and a positive input which
receives the second input line; a second invert adder having
another negative input which receives the third input line and
another positive input which receives the second input line; and a
bandpass filter which receives outputs from said first and second
invert adders and respectively generates the first and second
chrominance signals.
25. A video decoder comprising: one or more line stores which
produce a plurality of input lines from a composite video signal; a
subtractive comb and notch filter which receives said input lines
and generates a first pair of separated luminance and chrominance
signals; an additive comb filter which receives said input lines
and generates a second pair of separated luminance and chrominance
signals; a multiplexer which receives the first and second pairs of
separated luminance and chrominance signals and selectively outputs
one of said pairs; and an adaptive decision unit which receives
said input lines and controls said multiplexer based on any
luminance similarity among said input lines.
26. The video decoder of claim 25 wherein said adaptive decision
unit further enables notch filtering by said subtractive comb and
notch filter when said input lines have no luminance similarity and
no chrominance similarity.
27. The video decoder of claim 26 wherein said adaptive decision
unit receives chrominance signals from one of said subtractive comb
and notch filter or said additive comb filter.
28. The video decoder of claim 25 wherein: two line stores produce
first, second, and third input lines; and said adaptive decision
unit determines any luminance similarity among all three of said
first, second, and third input lines.
29. The video decoder of claim 28 wherein said adaptive decision
unit controls said multiplexer to pass said first pair of separated
luminance and chrominance signals when no luminance similarity
exists among all three of said first, second, and third input
lines, and to pass said second pair of separated luminance and
chrominance signals when some luminance similarity exists among all
three of said first, second, and third input lines.
30. The video decoder of claim 29 wherein said adaptive decision
unit further controls said additive comb filter to perform
three-line averaging when a high degree of similarity exists among
all three of said first, second, and third input lines.
31. An adaptive circuit for separating luminance and chrominance
components from a composite video signal comprising: a first input
line; two line stores connected to said first input line which
generate second and third input lines; a subtractive comb filter
which receives said first, second, and third input lines and
generates a first pair of separated luminance and chrominance
signals; an additive comb filter which receives said first, second,
and third input lines and generates a second pair of separated
luminance and chrominance signals; a multiplexer which receives the
first and second pairs of separated luminance and chrominance
signals and selectively outputs one of the first and second pairs;
and an adaptive decision unit which receives said first, second,
and third input lines and controls said multiplexer to pass said
first pair of separated luminance and chrominance signals when no
luminance similarity exists among all three of said first, second,
and third input lines, and to pass said second pair of separated
luminance and chrominance signals when some luminance similarity
exists among all three of said first, second, and third input
lines.
32. The adaptive circuit of claim 31 wherein said adaptive decision
unit further controls said additive comb filter to perform
three-line averaging when a high degree of similarity exists among
all three of said first, second, and third input lines.
33. The adaptive circuit of claim 31 wherein said adaptive decision
unit further determines any chrominance similarity among said
first, second, and third input lines and enables notch filtering by
said subtractive comb filter when said first, second, and third
input lines have no luminance similarity and no chrominance
similarity.
34. The adaptive circuit of claim 33 wherein said adaptive decision
unit receives chrominance signals from one of said subtractive comb
filter or said additive comb filter.
35. An adaptive decision unit for controlling an electronic video
decoder having a plurality of component filters which separate
luminance and chrominance components from a composite video signal,
the adaptive decision unit comprising: first, second, and third
input lines; a filter circuit which generates first and second
chrominance signals based on said first, second, and third input
lines; an adder circuit which generates first and second luminance
differences based on said first, second, and third input lines; and
a logic circuit which receives the first and second luminance
differences and the first and second chrominance signals and
activates one or more output signals to indicate that no luminance
similarity exists among the first, second, and third input lines
and that said first, second, and third lines have chrominance
similarity.
36. The adaptive decision unit of claim 35 wherein said filter
circuit includes: a first invert adder having a negative input
which receives the first input line and a positive input which
receives the second input line; a second invert adder having
another negative input which receives the third input line and
another positive input which receives the second input line; and a
bandpass filter which receives outputs from said first and second
invert adders and respectively generates the first and second
chrominance signals.
37. A method of separating luminance and chrominance components
from a composite video signal, comprising: deriving at least three
consecutive input lines from a composite video signal; determining
whether any luminance similarity exists among the three consecutive
input lines, and whether any chrominance similarity exists among
the three consecutive input lines by generating a first chrominance
value using a first one of the consecutive input lines and a second
one of the consecutive input lines, generating a second chrominance
value using the second consecutive input lines and a third one of
the consecutive input lines, computing a difference of the first
and second chrominance values, and comparing the difference to a
threshold value; and selectively enabling one of a plurality of
component filters based on said determining to separate the
luminance and chrominance components.
38. A method of separating luminance and chrominance components
from a composite video signal, comprising: deriving at least three
consecutive input lines from a composite video signal; determining
whether any similarity exists among the three consecutive input
lines; and selectively enabling one of a plurality of component
filters based on said determining to separate the luminance and
chrominance components, wherein a notch filter is enabled when no
luminance similarity exists between any of the three consecutive
input lines.
39. A method of separating luminance and chrominance components
from a composite video signal, comprising: deriving at least three
consecutive input lines from a composite video signal; determining
whether any similarity exists among the three consecutive input
lines; and selectively enabling one of a plurality of component
filters based on said determining to separate the luminance and
chrominance components, wherein a notch filter is enabled when no
chrominance similarity exists between the three consecutive input
lines.
Description
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention generally relates to electronic video
devices, and more particularly to a method of decoding or
separating video signal components (i.e., chrominance and
luminance) in a composite video signal.
2. Description of the Related Art
Various electronic video devices such as televisions, video
cassette recorders and digital video disc (DVD) players utilize a
composite video signal to record, transmit, and reproduce video
images. The composite video signal typically includes a luminance
(intensity) component denoted "Y" and a chrominance (color)
component denoted "C". This type of video signal encoding is common
to two well-known video standards, the National Television System
Committee (NTSC) standard used in North America and Japan, and the
Phase Alternating Line (PAL) standard used in most of Europe,
Africa and Asia. The chrominance component of the signal is encoded
on a subcarrier and added to the luminance signal. For NTSC, the
chrominance is modulated on a subcarrier frequency of 3.579545
megahertz (MHz), and for PAL the chrominance is modulated on a
subcarrier frequency of 4.433619 MHz, using quadrature
modulation.
When the transmitted composite video signal is received at a video
device, the luminance and chrominance components need to be
separated out in order to determine their respective values and
effectuate the color scheme. Many different video decoder designs
have been devised for this purpose. Some of these decoders use one
or more line delays and adders to cancel out the luminance
component, and a bandpass filter to obtain the chrominance
component. A conventional two-line delay YC separator circuit 10
for general viewing of a composite NTSC video signal is depicted in
FIG. 1.
The composite video signal input of separator circuit 10 is fed to
a first line delay 12 and an adder 14. The output of first line
delay 12 passes to a second line delay 16 and two invert adders 18
and 20. The output of second line delay 16 passes to the other
input of adder 14. Adder 14 generates a double-amplitude composite
video signal since the subcarriers are in-phase. A divider 22
(i.e., 0.5 multiplier) is used to normalize the signal, which is
then fed to the negative input of invert adder 18. Since a
180.degree. phase difference exists between the output of adder 14
and the one line-delayed composite video signal, most of the
luminance is canceled by adder 18, leaving double-amplitude
chrominance. Another divider 24 is used to normalize the
chrominance signal, which is then fed to a bandpass filter 26. The
output of bandpass filter 26 is the chrominance output signal. This
signal is also fed to the negative input of adder 20 to yield the
luminance output signal.
Many composite video signal decoders use comb filters. Comb filters
combine a scan line with one or more previous scan lines (scan
lines are a horizontal line as displayed on a monitor or screen).
Under the NTSC format, the chrominance phase changes 180.degree.
from one scan line to the next scan line. As a result, if two
adjacent scan lines are identical, then adding them will eliminate
the chrominance component, leaving only luminance. The same concept
can be applied to PAL signals by using four line delays.
Several disadvantages to line-delay circuits, such as that shown in
FIG. 1 exist. A principal disadvantage is the unsuppressed
cross-luminance on vertical color transitions, i.e., incorrect
decoding of the luminance value due to abrupt color changes.
Circuits that use comb filters can have further problems with
diagonal lines and vertical color changes since only
vertically-aligned samples are processed. Also, with diagonal
lines, after standard comb filtering, the chrominance information
includes the difference between adjacent luminance values, which
may be misinterpreted as chrominance information. This difference
can appear as cross-color artifacts along the edge of the line.
Sharp vertical color transitions can further generate the "hanging
dot" pattern commonly seen on the scan line between the two color
changes.
Several different adaptive decoders have been designed which
attempt to resolve these problems in Y-C separation. For example,
U.S. Pat. No. 6,462,790 discloses a digital comb filter for
decoding composite video signals which uses a fast Fourier
transform circuit or band split filter circuit to determine
characteristics of the input video signal without demodulating the
signal. Those circuits produce signature signals, which are then
used to correlate each of the video lines and compute weighting
coefficients for a comb filter. The device uses the sum of four
surrounding lines that are out of phase with the current line for
comb filtering. The comb filter reverts to a band split filter if
none of the surrounding lines are similar. The weighting
coefficients can be adjusted as the noise level increases or
decreases.
U.S. Patent Application Pub. No. 2002/0140866 discusses an adaptive
comb filter design for separating chrominance and luminance
components, which provides a threshold determination of whether the
lines of chrominance are correlated, and uses a comb filter to
separate out the Y-C information. If adjacent scan lines of
chrominance contain the same color information or differ by only a
slight level, the scan lines are considered correlated. If the
lines are uncorrelated, then a bandpass filter is used to separate
out the chrominance component from the composite video signal.
U.S. Patent Application Pub. No. 2002/0149702 describes another
decoder for composite video signals, which prevents decreases in
resolution at the time of Y-C separation associated with images
having vertical stripes. A correlation judging section determines
whether there is any correlation among neighboring signal lines,
and one of two filters (either a two-line comb filter or a
three-line comb filter) is selectively used depending on the
correlation. A stripe component judging section examines
neighboring pixels on the signal to indicate the presence of a
stripe.
U.S. Patent Application Pub. No. 2003/0071921 teaches a
luminance-chrominance signal separation device, which detects
diagonal components of the luminance signal and then utilizes
different bandpass filters to separate out the chrominance signal
based on the amount of diagonal components in the luminance signal.
Use of two different bandpass filters (a broad-bandpass filter and
a narrow-bandpass filter) reduces cross-color in the output
chrominance signal and improves resolution in the diagonal
direction of the output luminance signal.
While each of these designs has certain advantages, they all still
suffer from various limitations that are not addressed simply by
the selective use of bandpass filters versus comb filters. It
would, therefore, be desirable to devise an improved decoder for
separating out luminance and chrominance values in a composite
video signal, which more comprehensively addresses issues of
vertical luminance resolution, cross-chrominance artifacts, and
chrominance transitions. It would be further advantageous if the
decoder could retain high quality transmission of signals with no
chrominance content.
SUMMARY OF THE INVENTION
One object of the present invention is to provide an improved
decoder for separating luminance and chrominance components in a
composite video signal.
Another object of the present invention is to provide such a
decoder which adaptively processes the composite video signal to
minimize or reduce decoding errors from cross-chrominance artifacts
or vertical luminance transitions.
Yet another object of the present invention is to provide a
composite video signal decoder, which provides increased resolution
of the luminance component.
The foregoing objects are achieved in an adaptive circuit and
method which separates luminance and chrominance components from a
composite video signal by deriving three input lines from the
composite video signal, determining whether any similarities among
the three input lines exist, and then selectively enabling one of a
plurality of component filters based on any similarities. More
particularly, the luminance similarity is determined by generating
first and second luminance differences using different pairs of the
three input lines, and comparing each of the luminance differences
to a threshold value. If no luminance similarity exists among all
three of the input lines, a subtractive comb filter is enabled to
maintain high vertical luminance resolution. If luminance
similarity does exist among all three of the input lines, an
additive comb filter is enabled. The additive comb filter performs
three-line averaging when a high degree of similarity exists among
all three consecutive input lines to minimize cross-chroma
artifacts on lines that are similar.
Chrominance similarity among the three input lines can also be
determined by generating first and second chrominance values using
different pairs of the three input lines, computing a difference of
the first and second chrominance values, and comparing the chroma
difference to a threshold value. If no luminance similarity exists
between any of the three input lines or no chrominance similarity
exists (i.e., vertical chroma transition exists), then a notch
filter incorporated into the subtractive comb filter can be
enabled.
The above as well as additional objectives, features, and
advantages of the present invention will become apparent in the
following detailed written description.
BRIEF DESCRIPTION OF THE DRAWINGS
The present invention may be better understood, and its numerous
objects, features, and advantages made apparent to those skilled in
the art by referencing the accompanying drawings.
FIG. 1 is a high-level schematic diagram of a conventional two-line
delay Y-C separator circuit for general viewing of a composite
National Television System Committee (NTSC) video signal having
luminance ("Y") and chrominance ("C") components;
FIG. 2 is a high-level schematic diagram of one embodiment of an
adaptive, three-line delay circuit for Y-C separation of an NTSC
composite video signal constructed in accordance with the present
invention;
FIG. 3 is a schematic diagram illustrating the inputs to the
adaptive decision logic used in the adaptive circuit of FIG. 2;
FIG. 4 is a chart illustrating the decision flow used by the
adaptive decision logic of FIG. 3;
FIG. 5 is a schematic diagram showing one embodiment of the
subtractive comb and notch filter used in the adaptive circuit of
FIG. 2; and
FIG. 6 is a schematic diagram depicting one embodiment of the
additive comb filter used in the adaptive circuit of FIG. 2.
The use of the same reference symbols in different drawings
indicates similar or identical items.
DESCRIPTION OF THE PREFERRED EMBODIMENT(S)
With reference now to FIG. 2, one embodiment 30 of an adaptive
circuit for Y-C separation of an NTSC composite video signal
constructed in accordance with the present invention is depicted.
Circuit 30 generally includes two delays or line stores 32a and
32b, an adaptive decision unit 34, a three-line subtractive comb
and notch filter 36, a three-line additive comb filter 38, and a
multiplexer 40. In the illustrative embodiment, circuit 30 is
adapted to receive a digitized input of the composite video signal
from a 10-bit analog-to-digital converter (not shown) as part of a
larger video system, such as a digital video disc (DVD) player. The
values in the digitized input stream are accordingly in the range
of 0 to 1023. The three input lines L1, L2 and L3 are derived from
consecutive scan lines using the two line stores 32a and 32b. Each
line has 1440 samples. While this embodiment is designed for use in
an NTSC system, those skilled in the art will appreciate that the
invention is not so limited and could be applied to other systems,
such as PAL. For use in a PAL system, the adaptive circuit could
utilize four lines (three line stores) instead of three lines (two
line stores).
As explained in further detail below, the behavior of circuit 30
adapts based on the content of the three input lines to yield high
vertical luminance resolution, low cross-chrominance artifacts, and
excellent transmission quality for signals with no chrominance
content. The three input lines L1, L2 and L3 are passed to adaptive
decision unit 34 as well as filters 36 and 38. Adaptive decision
unit 34 generates five chrominance select (CSEL) signals and a
notch filter (NF) signal which are used to control filters 36 and
38 and multiplexer 40. The NF and CSEL[0,4] signals control
subtractive comb and notch filter 36, while the CSEL[1,2] signals
control additive comb filter 38. The CSEL[3] signal controls
multiplexer 40. Each of the filters 36, 38 has a pair of outputs
representing the separated luminance (Y) and chrominance (C)
components, and these component pairs are inputs to multiplexer 40,
which merely acts to select one of these pairs based on the setting
of CSEL[3]. Thus, the output of multiplexer 40 is the output of
adaptive circuit 30, i.e., the decoded Y and C signals.
Adaptive decision unit 34 is shown in greater detail in FIG. 3. The
input lines L1, L2, L3 pass through respective notch filters 42a,
42b, 42c. Each of the notch filters has a notch frequency of around
3.58 MHz to eliminate the chrominance component. The notch-filtered
luminance of input line L1 is connected to the negative input of an
invert adder 44a, and the notch-filtered luminance of input line L2
is connected to the positive input of invert adder 44a. The output
of invert adder 44a generates a first luminance difference
.DELTA.Y.sub.1 (luminance of input line L2--luminance of input line
L1). Similarly, the notch-filtered luminance of input line L3 is
connected to the negative input of another invert adder 44b, and
the notch-filtered luminance of input line L2 is connected to the
positive input of invert adder 44b. The output of invert adder 44b
generates a second luminance difference .DELTA.Y.sub.2 (luminance
of input line L2--luminance of input line L3). Low-pass and
adaptive logic 46 uses the luminance differences to determine the
similarity between the luminance content of the three lines. If
both of the differences are small (less than a threshold amount),
then a high-degree of similarity exists between all three lines,
and low-pass and adaptive logic 46 sets the NF and CSEL[ ] signals
to enable use of additive comb filter 38. The luminance difference
threshold amount may be hard wired but is preferably programmably
set using another input (comb_thresh) to low-pass and adaptive
logic 46. An exemplary threshold for luminance difference is 64
(decimal), an empirical value that represents a satisfactory cutoff
for discriminating between similar versus dissimilar luminances. If
either of the luminance differences are above the threshold, then
they are further compared, and low-pass and adaptive logic 46 sets
the NF and CSEL[ ] signals to enable use of subtractive comb filter
36, utilizing the lower difference of the two.
The similarity between chrominance components is also examined by
low-pass and adaptive logic 46 to detect vertical chrominance
transitions over three lines. First and second chrominance values
chromal2_bp and chroma23_bp are generated by feeding first and
second luminance differences .DELTA.Y.sub.1 and .DELTA.Y.sub.2
through a bandpass filter, which is discussed further below in
conjunction with FIG. 5. If the difference between these values is
more than a threshold value, then the NF signal is activated to
select use of the notch filter. The same threshold value
(comb_thresh) used for luminance similarity can be used for
detecting vertical color transitions.
Circuit 30 can further be forced to use the notch filter regardless
of any luminance or chrominance similarity by programmably setting
an enable input signal notch_en. The extent of adaptiveness of the
circuit (its resolution) can additionally be controlled using the
"adaptive_res" signal.
The operation of low-pass and adaptive logic 46 may be further
understood with reference to the decision flowchart of FIG. 4. The
process begins by initializing the control signals (CSEL, NF) to
zero at block 50, and first determining whether a notch filter (in
three-line subtractive comb and notch filter 36) should be used at
block 51. The notch filter (92, see FIG. 5) is used when no
luminance similarity exists (i.e., .DELTA.Y.sub.1>comb_thresh
and .DELTA.Y.sub.2>comb_thresh) and no chrominance match exists
(i.e., chromal2_bp-chroma23_bp<comb_thresh), or when the
notch_en signal is active. If either of these conditions is
satisfied, then the NF signal is turned on at block 52 (i.e.,
NF=1). If the notch filter is not to be used, then the next step
from decision block 51 is to decide whether to use the additive
filter 38 or the subtractive filter 36 by determining whether some
luminance similarity exists between all three input lines L1, L2
and L3 at decision block 54. If luminance similarity exists between
all three lines (i.e., .DELTA.Y.sub.1<comb_thresh/2 and
.DELTA.Y.sub.2<comb_thresh/2), then the additive filter 38 is
used, but if no similarity exists between all three lines, then the
subtractive filter 36 is used instead.
If the additive filter 38 is to be used, then the next step from
block 54 is to decide which set of lines to use for additive comb
filtering by determining whether a high degree of luminance
similarity exists between all three input lines at decision block
56. If a high degree of luminance similarity exists between all
three lines (i.e., .DELTA.Y.sub.1<comb_thresh/4 and
.DELTA.Y.sub.2<comb_thresh/4), then all three lines are used
with additive comb filter 38 by setting CSEL[1]=1 at block 58. A
three-line averager is used to minimize cross-chroma artifacts on
similar lines. If a high degree of luminance similarity does not
exist between all three lines, then the next step from decision
block 56 is to decide whether a high degree of luminance similarity
exists between at least two lines at decision block 60. If a high
degree of luminance similarity exists between lines L1 and L2
(i.e., .DELTA.Y.sub.1<comb_thresh/4), then CSEL[2] is set to 1
at block 62. If no high degree of luminance similarity exists
between any of the lines, then CSEL[2] is set to 0 at block 64. For
all cases that use the additive comb filter, CSEL[3] is also set to
1 at block 66.
Returning to decision block 54, if some luminance similarity does
not exist across all three lines, then the subtractive comb
datapath is used to maintain high vertical luminance resolution.
Low-pass and adaptive logic 46 makes a determination of whether a
high chrominance similarity exists across all three lines at
decision block 67. If high chrominance similarity does exist (i.e.,
chromal2_bp-chroma23_bp<comb_thresh/2), then CSEL[4] is set to 1
at block 68. Low-pass and adaptive logic 46 makes a further
determination of whether a high luminance similarity exists between
at least lines L1 and L2 at decision block 69. If such a
determination is made (i.e., .DELTA.Y.sub.1<comb_thresh/2), then
CSEL[0] is set to zero at block 70; otherwise CSEL[0] is set to 1
at block 72.
Subtractive comb and notch filter 36 is shown in further detail in
FIG. 5. Input line L1 is connected to the negative input of an
invert adder 80a, and input line L2 is connected to the positive
input of invert adder 80a. Similarly, input line L3 is connected to
the negative input of another invert adder 80b, and input line L2
is connected to the positive input of invert adder 80b. The outputs
of adders 80a and 80b are then fed to a bandpass (BP) filter 82.
The respective outputs of bandpass filter 82 form the chromal2_bp
and chroma23_bp signals that are used by low-pass and adaptive
logic 46. Those signals are also inputs to an adder 83 and to a
multiplexer 84 which is controlled by CSEL[0]. The outputs of adder
83 and multiplexer 84 are inputs to another multiplexer 85 which is
controlled by CSEL[4]. The output of multiplexer 85 is connected to
a divider 86, which divides the doubled chrominance signals by two
to derive the chrominance component. The output of divider 86 is
further connected to one input of another multiplexer 88, which is
controlled by the NF signal from low-pass and adaptive logic
46.
The output C_out of multiplexer 88 (the chrominance component) is
connected to the negative input of another invert adder 80c, and
input line L2 is connected to the positive input of invert adder
80c. The resulting luminance signal is connected to one input of
another multiplexer 90, which is also controlled by the NF signal.
The input line L2 is also connected to a notch filter 92, which
passes the notch-filtered luminance signal to the other input of
multiplexer 90, and to the negative input of another invert adder
80d. The original input line L2 is connected to the positive input
of invert adder 80d, and the resulting output of invert adder 80d
is only the chrominance component (since the luminance has been
subtracted out), which is connected to the other input of
multiplexer 88.
Additive comb filter 38 is shown in further detail in FIG. 6. Input
line L1 is connected to one (positive) input of an adder 94a, and
input line L2 is connected to the other (positive) input of adder
94a. Similarly, input line L3 is connected to one (positive) input
of another adder 94b, and input line L2 is connected to the other
(positive) input of adder 94b. The outputs of adders 94a and 94b
are connected to respective inputs of a multiplexer 96, which is
controlled by the CSEL[1] signal. The output of multiplexer 96
passes to a divider 98, which divides the doubled luminance signal
by two. The output of divider 98 is further connected to one input
of another multiplexer 100, which is controlled by the CSEL[2]
signal. The output Y_out of multiplexer 100 is connected to the
negative input of an invert adder 102, and input line L2 is
connected to the positive input of invert adder 102. The resulting
output of invert adder 102 is the chrominance signal (since the
luminance Y_out was subtracted out). Input line L1 is also
connected to one input of another adder 94c, and input line L3 is
connected to the other input of adder 94c. The output of adder 94c
is coupled to the other input of multiplexer 100 via another
divider 99 which divides the luminance signal by 4 (0.25.times.
gain).
Although the invention has been described with reference to
specific embodiments, this description is not meant to be construed
in a limiting sense. Various modifications of the disclosed
embodiments, as well as alternative embodiments of the invention,
will become apparent to persons skilled in the art upon reference
to the description of the invention. It is therefore contemplated
that such modifications can be made without departing from the
spirit or scope of the present invention as defined in the appended
claims.
* * * * *